An Observational Study of Hospital Paging Practices and Workflow Interruption Among On-Call Junior Neurological Surgery Residents

Abstract

Background

Previous studies suggest that nonurgent pages comprise a substantial portion of the pages received by residents while on duty. We evaluated the number, type, and urgency of pages received and the task being performed at the time of paging by on-call junior neurosurgery residents at a large teaching hospital, with the aim of providing insight into mechanisms that can be developed to improve paging patterns and ultimately reduce physician distractions due to nonurgent communications.

Methods

For eight 12-hour call sessions, a medical student shadowed the on-call junior neurosurgery resident and recorded all pages received and the time, paging number and location, priority of the page (nonurgent, urgent, or emergency), and the activity the resident performed when the page was received. During one 5-hour session, a recorder measured the amount of time spent returning pages.

Results

During the study period, 439 communications were recorded (mean of 54.9 per 12-hour session; range, 33–75). Communications occurred at a rate of every 13 minutes and ranged from every 34 minutes to every 8.7 minutes. Paging remained frequent even during the hours when on-call residents are most likely to sleep (2–5 am), with an average of 4 communications per hour. The time to return pages ranged from 15 to 174 seconds (mean, 79.7 seconds). Most pages were nonurgent (68.3%) and occurred during patient care activities (65%).

Conclusions

Paging communications were frequent. Most pages were nonurgent and were received during important patient care activities. This suggests that a viable solution must address the work context of the individual being paged and the individual initiating the page to ensure that urgent communications are properly prioritized and attended to.

Introduction

The purpose of pager communication in a hospital environment is to relay important information between health care providers. Previous studies suggest that nonurgent items make up a substantial portion of the pages received by on-duty residents.^1–3 This may distract them from ongoing tasks and emergency activities, and it may be a potential cause of medical errors. At the same time, some important information is not relayed to on-call physicians to avoid bothering the doctor, and this may result in the delay of necessary diagnosis or treatments.

Paging practices and their burden on on-call residents have been studied in a variety of settings.^1–9 To date, only one study, conducted 2 decades ago, prospectively collected page priorities and the activities performed by the resident at the time pages were received.² Few studies have evaluated paging practices in surgical specialties, and none have prospectively evaluated paging priorities and the interruption of resident activities since the onset of the current duty hour limits.

The purpose of this study is to evaluate the number, type, and urgency of pages received by on-call junior neurosurgery residents at a large teaching hospital, as well as the tasks that were performed at the time the resident was being paged. The aim is to provide insight into how paging patterns could be improved to reduce physician distractions that result from nonurgent communications.

Methods

Setting

The University of Florida neurological surgery residency program has 3 residents annually, and the department performs approximately 5000 operations per year. While on call, residents are responsible for covering 2 geographically proximate hospitals. A junior resident (postgraduate year 2–3 [PGY-2–3]) and a senior resident (PGY-4–6) are on call during all overnight periods. Junior and senior resident call lasts for a period of 24 hours. The on-call attending does not take in-house call but is available. The junior resident carries 2 hospital pagers and receives most pages from nurses and physicians, as well as patient phone calls. The junior resident evaluates new consults, manages hospital-to-hospital transfers, and attends to inpatient events (eg, patients with neurological decline, falls, etc), and also performs minor procedures. The senior resident helps with work as needed and assists the on-call attending in the operating room should an emergency operation be necessary.

Study Design

For 8 call sessions, a medical student recorder with Health Insurance Portability and Accountability Act compliance training shadowed the on-call neurosurgical junior resident during call periods from 7 pm to 7 am on Friday and Saturday nights from October through November 2011. Seven junior neurosurgical residents (of 19 total departmental residents) were eligible to participate. The week prior to study initiation, 2 recorders and the 7 junior residents were oriented to the study goals and the specifics of what was to be recorded during each call period. Each time a page was received by the on-call junior resident, the recorder logged the time, paging number and location, priority of the page (nonurgent, urgent, or emergency, based on resident input), and the activity that was being performed when the page was received. Additionally, “walk-ups” (ie, when health care providers approach the resident on call with patient care issues without paging) were recorded. Activities were categorized into seeing a new patient, patient care activity, procedure, sleep or meal, and other low urgency (box). Furthermore, during one 5-hour session, a recorder measured the amount of time spent returning individual pages. Start time was defined as the moment the resident picked up the telephone to dial and stop time as the moment the phone was hung up. If the resident could not reach the intended individual (busy or did not answer), times from all subsequent attempts were added to the first to determine a total time for that given page. Finally, repeat pages were defined as 2 or more pages with essentially the same message and to the same call-back phone number that occurred before the first page could be returned.

What was known

Duty hour limits compress residents' time, and data suggest nonurgent pages comprise a substantial portion of pages received, and may occupy a significant portion of residents' time.

What is new

Research showed that most pages to junior neurosurgical residents were nonurgent and many were received during patient care activities, potentially distracting residents from important tasks.

Limitations

Small sample and a single setting reduces the ability to generalize; results may be biased by residents' subjective assessment of the urgency of pager communications.

Bottom line

Solutions need to address the work context of the individuals being paged and those initiating the page to ensure urgent communications are prioritized and attended to.

Box Examples of Page Priorities and Resident Activities

University of Florida Institutional Review Board approval was obtained prior to study initiation. Data analysis was performed using SPSS version 16.0 (IBM Corp, Armonk, NY).

Results

During the study period (96 hours), 439 pager communications were recorded (mean of 54.9 per 12-hour shift; median, 56; range, 33–75). Of these, 405 (92.3%) were pages and 34 (7.7%) were walk-up communications. Each of the 7 eligible junior residents participated during at least 1 call period. Pages were more frequent from 7 pm to midnight (figure). On average, communications occurred at a rate of 4.6 per hour, or 1 communication every 13 minutes, and ranged from as low as 1.8 per hour from 5 to 6 am (every 34 minutes) to as high as 6.9 per hour from 7 to 8 pm (every 8.7 minutes). Paging remained frequent from 2 to 5 am, when on-call residents are most likely to sleep, with an average of 4 communications per hour (1 communication every 15 minutes); during this time period 96 total communications occurred, with most being nonurgent (n  =  68; 70.8%).

Figure.

Total Pages Received Per Half-Hour Time Interval for All 8 Call Shifts

During the 5-hour session where time spent returning pages was reported, 21 pages occurred. The time to return these pages ranged from 15 to 174 seconds (mean, 79.7 seconds).

Ninety-nine percent (435 of 439) of the communications had priority and location reported (table 1). Most communications originated from the intensive care unit (n  =  193; 44.6%) and were nonurgent (n  =  297; 68.3%), whereas most emergency or urgent pages originated from the emergency room.

Table 1.

The Total Number and Priority of Communications by Location of Origin

Resident activity at time of being paged was reported in 434 of 439 communications (98.9%; table 2). The most common activities were routine patient care activities (n  =  284; 65%) followed by new patient evaluations (n  =  86; 20%).

Table 2.

Activities Performed by the Junior Residents at Time of Communications

A total of 31 repeat pages regarding 27 communications occurred during the 8 call shifts (7.1% of total pages; 3.9 repeat pages per shift). The average time between the initial page and the repeat page was 9.0 minutes (range, 2–24 minutes; median, 7 minutes). All repeat pages occurred while the resident was performing a procedure, seeing a new patient, or engaged in patient care activities. Of the 31 repeat pages, 3 (9.7%) were emergency, 15 (48.4%) were urgent, and 13 (41.9%) were nonurgent. Repeat pages were more likely to be emergency or urgent than nonrepeat pages (P  =  .005).

Based on a mean time of 80 seconds per page and an average of approximately 50 pages per 12-hour shift, residents on call spent an estimated 67 minutes (1.1 hours) on the phone during each 12-hour period (9% of the on-call shift).

Discussion

Junior residents received a sizable number of pages, of which a significant percentage (68.3%) concerned nonurgent matters. Pages often were received while residents were actively involved in important patient care activities, and nearly 10% of junior residents' 12-hour shift was spent on the telephone returning pages. The volume of pages and their frequently nonurgent status may represent potential distractions for junior residents that may compromise the quality and safety of care, as has been suggested by other studies.⁴

Previous studies have revealed a high degree of satisfaction among physicians with alpha-numeric paging systems, which can minimize interruptions in patient care^5,6; however, a recent study questioned the utility of this approach, citing a high incidence of poor or inadequate communications.⁷ Two-way paging is another alternative and was preferred over one-way alpha-numeric pagers in one small study.⁸ Smart phones with e-mail capabilities have also been evaluated in several studies recently, with subjective improvements in communication between nurses and physicians, although e-mail messages share similar shortcomings to text paging.^10–12

One potential answer is to have nonurgent pages clustered among individual patient units so that a page is initiated only after a certain threshold number of communications are reached. In this model, individual nurse requests would be pooled, and once 3 or 4 nonurgent requests had accumulated, the resident on call would be paged. This would reduce the amount of time on the phone and the number of individual distractions. It would not interfere with paging for urgent or emergency issues. For example, if all nonurgent patient issues arising between 2 and 5 am were held until the morning (eg, if only urgent or emergency pages were initiated during these hours), pages would occur at a rate of only 1.2 per hour, or once every 51.7 minutes. This could potentially increase the amount and quality of sleep obtained by on-call residents. At the same time, this solution has the potential to postpone important communications to the on-call physicians because of an underestimated urgency by the individual who is clustering the communications.

Our study has several limitations. First, the paging practices experienced by our neurosurgical residents during this study may not necessarily be generalizable to other specialties or other institutions. Although an orientation session and handout were provided to instruct residents and recorders on how to prioritize pages, the results may be biased by individual resident fatigue, workload, or attitude because residents selected the page priority at the time of the page based on the individual situation (a blinded review process was not used). Walk-up communications were included because they represent important communications that demand resident time and represent potential distractions; however, these communications may not be an adequate surrogate for pages.

Conclusions

At our institution, pager communications were frequent and nonurgent, and most were received during important patient care activities. Batching of nonurgent pages may be a viable solution. Any solution must address the perspectives of both the person being paged and the person initiating the page to ensure the urgency of the communication is properly conveyed and addressed.